Method for estimating blending ratio of reconstituted tobacco in blended cut tobacco based on regression analysis

Abstract

The method for estimating the blending ratio of reconstituted tobacco in blended tobacco by regression analysis according to the present invention, wherein the blended tobacco includes pure tobacco and reconstituted tobacco, and the method includes: establishing a regression equation for blending reconstituted tobacco in the blended tobacco and determining the regression equation; calculating the blending ratio of reconstituted tobacco in the blended tobacco to be measured. It quantifies the actual blending ratio of reconstituted tobacco during the cigarette production process by establishing a relatively scientific model. Through practical verification, this method has relatively high accuracy, overcomes the inconvenience of manually sorting and calculating the blending ratio of reconstituted tobacco in the prior art, and facilitates the research and improvement of the utilization effect of reconstituted tobacco.

Description

Technical Field

[0001] The present invention relates to a method for estimating the blending ratio of reconstituted tobacco in blended tobacco by regression analysis, belonging to the technical field of tobacco. Background Art

[0002] Reconstituted tobacco is an essential raw material component in cigarette formulations, mainly made from tobacco fragments, tobacco dust, and tobacco stems through recombination and reconstruction. It plays an important role in reducing harm and tar, as well as reducing costs and increasing efficiency in cigarettes at home and abroad.

[0003] During the cigarette manufacturing process, reconstituted tobacco is blended and used in the form of leaf blending according to a certain proportion. Its blending ratio and blending uniformity have a significant impact on the precision control of cigarette formulations and the stability of sensory quality. However, since there is currently no fast and effective method or model in the industry to characterize and evaluate the actual blending ratio of reconstituted tobacco in finished tobacco and cigarettes, its actual blending ratio in the cigarette manufacturing process relies on manual sorting and calculation. Due to the significant subjective influence during the sorting process and the relatively fast moisture absorption characteristics of reconstituted tobacco, obvious moisture loss of reconstituted tobacco occurs during the sorting process, resulting in insufficient detection timeliness and accuracy of detection results. As a result, the processing process and technology optimization of reconstituted tobacco are restricted, and its utilization value cannot be maximized in the cigarette manufacturing process. Summary of the Invention

[0004] The purpose of the present invention is to provide a method for estimating the blending ratio of reconstituted tobacco in blended tobacco by regression analysis to solve the technical problems of insufficient detection timeliness and accuracy in the prior art.

[0005] To achieve the invention purpose of this application, the following technical solutions are adopted in this application:

[0006] A method for estimating the blending ratio of reconstituted tobacco in blended tobacco by regression analysis according to the present invention, the blended tobacco includes: pure tobacco and reconstituted tobacco, wherein: the method includes the following steps:

[0007] (1) Establish a regression equation for blending reconstituted tobacco in blended tobacco

[0008] (a) Select low environmental humidity and high environmental humidity

[0009] Mix the above-mentioned pure tobacco and the above-mentioned reconstituted tobacco in at least four proportions to obtain at least four samples. Divide each of the above samples into two equal parts. Under two different environmental humidities, perform moisture balance on each sample, and calculate the two-way moisture retention M of each sample. The calculation method is:

[0010] M = (Wmax - Wmin) / (Hhige - Hlow) Formula (1)

[0011] Wherein:

[0012] M is the two-way moisture retention of each blended cut tobacco at two different ambient humidities, Wmax is the dry basis equilibrium moisture content of each blended cut tobacco at high ambient humidity; Wmin is the dry basis equilibrium moisture content of each blended cut tobacco at low ambient humidity; Hhige is the high ambient humidity, and the high ambient humidity is 70%-90%; Hlow is the low ambient humidity, and the low ambient humidity is 30-50%; the two-way moisture retention M1, M2, M3, M4... of the blended cut tobacco containing at least four proportions of reconstituted tobacco are obtained;

[0013] (b) Obtain the regression equation and calculate the fitting determination coefficient R under the selected low and high ambient humidities above 2

[0014] Under the selected low and high ambient humidities above, fit the blending ratio of the reconstituted tobacco in the selected at least four blended cut tobaccos and their corresponding two-way moisture retention M1, M2, M3, M4... Through the regression equation: Φ = AM 2 + BM + C formula (2), obtain A, B, C and the fitting determination coefficient R 2 ,

[0015] Wherein: Φ is the blending ratio of the reconstituted tobacco in each blended cut tobacco, M1, M2, M3, M4... are the two-way moisture retention of their corresponding blended cut tobaccos. When the fitting determination coefficient R 2 <0.95, repeat step (a), re-select the low and high ambient humidities, and recalculate A, B, C in the regression equation and the fitting determination coefficient R 2 ; when the fitting determination coefficient R 2 > 0.95, obtain the determined low and high ambient humidities and the regression equation, and proceed to step (two);

[0016] (Two) Calculate the blending ratio of the reconstituted tobacco in the blended cut tobacco to be measured according to the determined regression equation

[0017] Divide the blended cut tobacco to be measured into two equal parts on average. Under the determined low and high ambient humidities, balance the moisture of the above two parts of the blended cut tobacco to be measured, and then calculate the two-way moisture retention M of each part of the blended cut tobacco to be measured according to formula (1) 0， Substitute M 0 into the determined regression equation to obtain the blending ratio Φ of the reconstituted tobacco in the blended cut tobacco to be measured 0 .

[0018] The method for estimating the blending ratio of reconstituted tobacco in blended cut tobacco based on regression analysis of the present invention, wherein: the blended cut tobacco to be measured needs to be balanced in an environment with a temperature of 22°C and a relative humidity of 60% for > 15 days.

[0019] The method for estimating the blending ratio of reconstituted tobacco in blended cut tobacco based on regression analysis according to the present invention, wherein: the blending ratio of reconstituted tobacco in the blended cut tobacco sample is 0%-25% of the weight of the blended cut tobacco.

[0020] The method for estimating the blending ratio of reconstituted tobacco in blended cut tobacco based on regression analysis according to the present invention, wherein: the temperatures under low environmental humidity and high environmental humidity are 25°C.

[0021] The method for estimating the blending ratio of reconstituted tobacco in blended cut tobacco based on regression analysis according to the present invention quantifies the actual blending ratio of reconstituted tobacco during the cigarette production process by establishing a relatively scientific model. Through practical verification, this method has high accuracy, overcomes the inconvenience of manually sorting and calculating the blending ratio of reconstituted tobacco in the prior art, facilitates the research and improvement of the utilization effect of reconstituted tobacco, and can quickly determine the content of reconstituted tobacco in the blended cut tobacco, and has simple operation and accurate measurement results. Description of the Drawings

[0022] Figure 1 It is a fitting curve graph in one embodiment for the method for estimating the blending ratio of reconstituted tobacco in blended cut tobacco based on regression analysis according to the present invention. Detailed Embodiments

[0023] To have a further understanding and recognition of the structural features and achieved effects of the present invention, the technical solutions of the present invention will be clearly and completely described below in conjunction with the embodiments.

[0024] The method for estimating the blending ratio of reconstituted tobacco in blended cut tobacco based on regression analysis according to the present invention, the blended cut tobacco includes: pure cut tobacco and reconstituted tobacco, and the method includes the following steps:

[0025] (1) Establish a regression equation for blending reconstituted tobacco in blended cut tobacco

[0026] (a) Select low environmental humidity and high environmental humidity

[0027] Select 1000 g each of pure cut tobacco of brand A and reconstituted tobacco of type B produced online in a certain cigarette factory. The pure cut tobacco is evenly divided into 6 parts. The reconstituted tobacco is evenly mixed with the cut tobacco at blending ratios of 0%, 5%, 10%, 15%, 20%, and 25% respectively to obtain six samples. Each of the above samples is evenly divided into two parts. Under two different environmental humidities, the moisture balance of each sample is carried out, and the two-way moisture retention M of each sample is calculated. The calculation method is:

[0028] M = (Wmax - Wmin) / (Hhige - Hlow) Formula (1)

[0029] Wherein:

[0030] M is the two-way moisture retention of each blended cut tobacco at two different ambient humidities, Wmax is the dry basis equilibrium moisture content of each blended cut tobacco at high ambient humidity; Wmin is the dry basis equilibrium moisture content of each blended cut tobacco at low ambient humidity; Hhige is the high ambient humidity, and the high ambient humidity is 70%-90%; Hlow is the low ambient humidity, and the low ambient humidity is 30-50%; the two-way moisture retention M1, M2, M3, M4, M5 and M6 of the blended cut tobacco containing at least four proportions of reconstituted tobacco are obtained;

[0031] (b) Obtain the regression equation and calculate the fitting determination coefficient R at the selected low ambient humidity and high ambient humidity 2

[0032] At the selected low ambient humidity and high ambient humidity, the temperature at low ambient humidity and high ambient humidity is 25°C. Fit the blending ratio of reconstituted tobacco in the six selected blended cut tobaccos and their corresponding two-way moisture retention M1, M2, M3, M4, M5 and M6. Through the regression equation: Φ = AM 2 + BM + C formula (2), obtain A, B, C and the fitting determination coefficient R 2 ,

[0033] Wherein: Φ is the blending ratio of reconstituted tobacco in each blended cut tobacco, M1, M2, M3, M4, M5 and M6 are the two-way moisture retention of their corresponding blended cut tobaccos. When the fitting determination coefficient R 2 <0.95, repeat step (a), reselect the low ambient humidity and high ambient humidity, and recalculate A, B, C and the fitting determination coefficient R in the regression equation 2 , and the pure tobacco and reconstituted tobacco used repeatedly are the same tobacco; when the fitting determination coefficient R 2 >0.95, as Figure 1 shown, obtain the determined low ambient humidity and high ambient humidity and the regression equation, low ambient humidity = 49.1%, high ambient humidity = 88.2%, and the regression equation is Φ = 513.46M 2 -1611.6M + 1252.6, R 2 = 0.9803, and proceed to step (two);

[0034] (Two) Calculate the proportion of reconstituted tobacco blended in the blended cut tobacco to be measured according to the determined regression equation

[0035] The blended cut tobacco to be measured needs to be balanced for more than 15 days in an environment with a temperature of 22°C and a relative humidity of 60%. The blended cut tobacco to be measured is evenly divided into two parts. Under the determined low environmental humidity and high environmental humidity, the moisture balance of the above two parts of the blended cut tobacco to be measured is carried out, and then the two-way moisture retention M of each part of the blended cut tobacco to be measured is calculated according to formula (1). 0， Substitute M 0 into the determined regression equation to obtain the blending ratio Φ of the reconstituted tobacco of the blended cut tobacco to be measured 0 .

[0036] Example 1

[0037] Select the above-mentioned pure cut tobacco of brand A and reconstituted tobacco of type B. The pure cut tobacco is blended with 20% of the reconstituted tobacco, and the measurement results obtained by manual sorting and the method of the present invention are compared.

[0038] 1. Select 1000 g of pure cut tobacco of brand A after baking and balanced moisture and 200 g of reconstituted tobacco of type B. Mix the above pure cut tobacco and reconstituted tobacco evenly, divide them into two parts on average, and stack them naturally and seal them for later use.

[0039] 2. For the first sample to be measured, use the method of manual sorting to detect that the blending ratio of its reconstituted tobacco is 18.6%.

[0040] 3. Evenly divide the second sample to be measured into two parts, measure the equilibrium moisture content under two different environmental humidities of low environmental humidity of 49.1% and high environmental humidity of 88.2%. According to M=(Wmax - Wmin) / (Hhige - Hlow) (formula 1), obtain the two-way moisture retention M of the second sample to be measured as M = 1.32. According to the determined regression equation: Φ = 513.46M 2 -1611.6M + 1252.6, obtain Φ = 19.94; all data are shown in Table 1, retaining two significant figures.

[0041] Hhige Hlow Wmax Wmin M Φ(%) 88.2％ 49.1％ 59.12％ 7.56％ 1.32 19.94

[0042] Table 1

[0043] If the theoretical blending ratio of 20% is taken as the true value, the prediction error between the result measured by the method of the present invention and the true value is 0.30%, and the absolute error is 0.06%; the prediction error between the result of manual detection and the true value is 7.0%, and the absolute error is 1.40%.

[0044] The results of Example 1 show that there is an obvious difference between the result of manual detection and the true value, and the blending ratio of the reconstituted tobacco calculated by this method is closer to the true value.

[0045] Example 2:

[0046] Select the above-mentioned pure tobacco cut filler of Brand A and reconstituted tobacco of Type B. The pure tobacco cut filler is blended with 30% of the reconstituted tobacco, and the measurement results obtained by manual sorting and the method of the present invention are compared.

[0047] 1. Select 1000 g of pure tobacco cut filler of Brand A after baking and 300 g of reconstituted tobacco of Type B. Mix the above-mentioned pure tobacco cut filler and reconstituted tobacco evenly, divide them into two equal parts, stack them naturally and seal for later use.

[0048] 2. For the first sample to be measured, use the method of manual sorting to detect that the blending ratio of the reconstituted tobacco is 28.63%.

[0049] 3. Divide the second sample to be measured evenly into two parts. Measure the equilibrium moisture content at two different ambient humidities of 49.1% (low ambient humidity) and 88.2% (high ambient humidity). According to the formula M = (Wmax - Wmin) / (Hhige - Hlow) (Formula 1), the two-way moisture retention of the second sample to be measured is M = 1.29. According to the determined regression equation: Φ = 513.46M 2 - 1611.6M + 1252.6, we get Φ = 28.08. All the data are shown in Table 2, retaining two significant figures.

[0050] Equilibrium moisture content on dry basis Hhige Hlow Wmax Wmin M Φ(%) 13.8％ 88.2％ 49.1％ 58.74 8.46 1.29 28.08

[0051] Table 2

[0052] If the theoretical blending ratio of 30% is taken as the true value, the prediction error between the result measured by the method of the present invention and the true value is 6.4%, and the absolute error is 1.92%; the prediction error between the result of manual detection and the true value is 4.57%, and the absolute error is 1.37%.

[0053] The results of Example 2 show that there is an obvious difference between the result of manual detection and the true value, and the blending ratio of the reconstituted tobacco calculated by the manual detection method is closer to the true value.

[0054] The results of the above two examples show that the result evaluated by the method of the present invention in Example 1 is significantly better than the result evaluated by the method of the present invention in Example 2. Thus, it can be seen that the method of the present invention can indirectly evaluate the blending ratio degree of the reconstituted tobacco. In Example 2, by increasing the blending ratio of the reconstituted tobacco, the prediction error increases significantly. When the blending ratio of the reconstituted tobacco is between 0 - 25%, the detection result of the method of the present invention is better than the result of manual sorting.

[0055] Compared with the prior art, the present invention has the following beneficial effects:

[0056] Through stepwise regression analysis of the bidirectional moisturizing property of reconstituted tobacco under different environmental humidities and decomposition of the correlation coefficients therein, the regression equation between the blending ratio and the bidirectional moisturizing property is obtained, providing a reference basis for the research on the utilization effect of reconstituted tobacco during the cigarette processing

[0057] By predicting the blending ratio of reconstituted tobacco through the optimal regression equation, the long process of traditional manual sorting can be avoided. Moreover, the prediction method is simple, fast, with high accuracy of the obtained blending ratio result, small absolute error, and less interference from subjective factors. It can facilitate the subsequent research on the processing technology and effect verification of reconstituted tobacco, and has very good application value.

[0058] The above is only the specific implementation manner of the present invention, but not a limitation to the present invention. Those skilled in the art can modify and improve it according to the basic idea of the present invention. However, all other embodiments obtained without creative labor fall within the scope of protection of the present invention.

Claims

1. A method for estimating the blending ratio of reconstituted tobacco in blended tobacco by regression analysis, the blended tobacco comprises: pure tobacco and reconstituted tobacco, characterized in that: the method comprises the following steps: (1). Establish a regression equation for the blended reconstituted tobacco in the blended tobacco (a). Select low environmental humidity and high environmental humidity Mix the above-mentioned pure tobacco and the above-mentioned reconstituted tobacco in at least four blending ratios to obtain at least four samples. Divide each sample into two equal parts. At two different environmental humidities, perform moisture balance on each sample, and calculate the two-way moisture retention M of each sample. The calculation method is: M = (W max - W min ) / (H hige - H low ) Equation (1) Wherein: M is the two-way moisture retention of each blended cut tobacco at two different ambient humidities, W max is the moisture content at dry basis equilibrium of each blended cut tobacco at high ambient humidity; W min is the moisture content at dry basis equilibrium of each blended cut tobacco at low ambient humidity; H hige is the high ambient humidity, and the high ambient humidity is 70%-90%; H low is the low ambient humidity, and the low ambient humidity is 30-50%; The two-way moisture retentions M1, M2, M3, M4... of the blended cut tobacco with at least the above four blending ratios are obtained; (b), under the selected low ambient humidity and high ambient humidity, obtain the regression equation and calculate the fitting determination coefficient R 2 Under the above selected low environmental humidity and high environmental humidity, the blending ratios of reconstituted tobacco in the above at least four selected blended cut tobaccos and their corresponding two-way moisturizing properties M1, M2, M3, M4... are fitted. Through the regression equation: Φ = AM 2 + BM + C formula (2), A, B, C and the fitting determination coefficient R 2 , Wherein: Φ is the blending ratio of reconstituted tobacco in each blended tobacco, M1, M2, M3, M4... are the two-way moisture retention properties of their corresponding blended tobaccos. When the fitting determination coefficient R 2 < 0.95, repeat step (a), reselect the low environmental humidity and high environmental humidity, and recalculate A, B, C in the regression equation and the fitting determination coefficient R 2 ; when the fitting determination coefficient R 2 > 0.95, obtain the determined low environmental humidity, high environmental humidity and regression equation, and proceed to step (two); (2). According to the determined regression equation, estimate the blending ratio of reconstituted tobacco in the blended tobacco to be measured The average of the blended cut tobacco to be measured is evenly divided into two parts. Under the determined low environmental humidity and high environmental humidity, the moisture balance of the above two parts of the blended cut tobacco to be measured is carried out, and then the two-way moisture retention M of the blended cut tobacco to be measured is calculated according to formula (1). 0 , substitute M 0 into the determined regression equation to obtain the blending ratio Φ of the reconstituted tobacco of the blended cut tobacco to be measured. 0 .

2. The method for estimating the blending ratio of reconstituted tobacco in blended tobacco by regression analysis according to claim 1, characterized in that: The blended tobacco to be measured needs to be balanced for > 15 days in an environment with a temperature of 22°C and a relative humidity of 60%.

3. The method for estimating the blending ratio of reconstituted tobacco in blended tobacco by regression analysis according to claim 2, characterized in that: The blending ratio of reconstituted tobacco in the sample of the blended tobacco is 0%-25% of the weight of the blended tobacco.

4. The method for estimating the blending ratio of reconstituted tobacco in blended tobacco by regression analysis according to claim 3, characterized in that: The temperature under the low environmental humidity and the high environmental humidity is 25°C.

Images